Volatility Based Tests for Informational Efficiency on Commodity Options Markets

This paper has two objectives. The first is to develop a simple, computationally tractable procedure for estimating implied GARCH volatilities from commodity options price data. The second is to apply this procedure to elicit implied volatilities from soybean option price data and investigate how well the resulting volatility forecasts predict ex-post "realized" volatilities. We find that filtering option prices through a GARCH option pricing model provides informative forecasts of daily volatilities, but that these forecasts can generally be improved upon using additional information available at the time the options are being priced. The results have implications for forecasting volatility, as well as for the informational efficiency of soybean options markets